3 Player Spades

Fifty-two divided by three is not a whole number, so something has to give. The near-universal solution is to remove one low card before the deal so that fifty-one cards go out seventeen at a time. The two of clubs is the conventional choice, though the two of diamonds and the two of hearts are both used and neither is wrong.

The one rule that is not negotiable is that you never remove a spade. Spades are trump, and pulling one out changes the trump count from thirteen to twelve, which quietly rebalances the entire game. Strip from a side suit and the trump structure stays intact.

The other accepted method is to deal seventeen cards each from the full deck and set the last card aside face down as a dead card. This is mathematically identical to stripping the two of clubs except that nobody knows which card is missing, which makes counting slightly harder and is the reason some players prefer it. A few tables give that last card to the winner of the first trick, but this leaves one player holding an extra card and creates more problems than it solves.

Fifty-one cards, three players, seventeen cards each, seventeen tricks per hand. Spades are trump. You must follow suit if you can. Nobody may lead a spade until spades have been broken by a ruff or until a player holds nothing else.

Bidding starts with the player to the dealer's left and moves clockwise. Every player bids for themselves and the three bids do not have to add up to seventeen. There are no partners, no team totals and no shared responsibility for anything.

Scoring is standard: make your bid and score ten points a trick plus one point for each overtrick, miss it and lose ten points a trick bid. Nil is normally worth one hundred and costs one hundred when broken. Ten accumulated bags cost one hundred points, and unlike the partnership game those bags belong to you alone.

Five hundred is the usual target. Because each hand contains seventeen tricks rather than thirteen, hands take longer and score higher, so plenty of groups drop the target to three hundred for a session that finishes in reasonable time.

Seventeen tricks divided three ways is a shade under six, so an average hand is a five or six bid rather than the three you would call at a four-player table. Players arriving from partnership Spades almost always underbid their first few three-handed hands by two or three tricks.

Trump counts run higher too. Thirteen spades across three hands averages out at four and a third each, against three and a quarter in the four-player game. Holding five spades is ordinary here and holding seven is a hand you can build a big bid around, because the remaining six are split between only two opponents.

Long side suits are the other thing that changes. With seventeen cards you regularly hold six or seven of something. If you have seven hearts, the other six are divided between two hands, so your fourth heart is often already a winner and the suit runs for several tricks before anyone can ruff. In four-player Spades that same seven-card suit dies much sooner. Count long-suit length as real tricks here, not as hope.

Nil is markedly harder. You have seventeen cards to keep clean instead of thirteen, and both opponents profit from busting you with nobody obliged to protect you. Some tables raise nil to a hundred and fifty to compensate for the difficulty. Most simply leave it at a hundred and see very few successful ones.

There are no partners, but there are constant temporary alliances, and they form without anyone saying a word. If you bid seven and both opponents bid four, they have a shared interest in setting you and no reason at all to protect each other's overtricks. Two players quietly working against one is the normal state of a three-handed hand.

The scoreboard drives it. Once somebody crosses four hundred, the other two start playing to set that player rather than to maximise their own score, because second place pays nothing. This is not collusion, it is just correct play, and you should expect it rather than complain about it.

The strategic response is to bid in a way that makes setting you expensive for the people trying. A bid of eight invites the whole table to spend the hand attacking you. A bid of five with a hand that can take seven collects two bags, keeps you off the radar, and lets somebody else absorb the pressure. Many strong three-handed players deliberately sit in second place until the last two hands and then convert.

The reverse also holds. When you are in last place nobody bothers to set you, which makes a big bid from behind far safer than the same bid from in front. Aggression is cheap when you are not the threat.

Every trick is contested by three cards instead of four, so second-best cards win far more often. Kings lead and hold. Queens survive. The general effect is that your middle cards are worth more than your partnership instincts suggest.

Counting is genuinely easy here, and it is the biggest edge available. You see seventeen of the fifty-one cards, which leaves thirty-four unknown cards split between two hands. Once a player shows out of a suit, you know the exact location of every remaining card in it. Two hands is a workload you can actually carry through seventeen tricks.

Trump management is the whole endgame. With everybody holding four or five spades, two rounds of trump leads strip most of the table, and whoever still has spades at trick fourteen takes the rest. The last four or five tricks of a three-handed hand are almost always pure trump, so budget for that and do not spend your last spade winning a trick you had already bid for.

If you stripped the two of clubs, clubs is a twelve-card suit and voids appear there marginally sooner. Leading clubs to force a ruff out of an opponent who is short is a small but real edge, and one that only exists because of how the deck was adjusted.

Three is the most common awkward number in card games, and the alternatives are worse than they look. Adding a computer player to make a four-handed partnership game means one human is stuck with a partner they cannot talk to and whose bidding they cannot influence, which is the single least enjoyable seat in Spades.

Three-handed keeps all three humans playing every hand for their own score. Nobody is carried, nobody is blamed, and the two-on-one dynamic gives it a negotiating texture the partnership game does not have.

The cost is time. Seventeen tricks a hand is a third longer than thirteen, so a game to five hundred is a real commitment. Drop the target if the session is short, and agree before the first deal on which card you are stripping so that nobody discovers it mid-hand.